Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
© Frontier Economics Pty. Ltd., Australia.
Assessment of ARTC’s Revenue
Allocation Methodology A REPORT PREPARED FOR HUNTER VALLEY ENERGY COAL
September 2014
i Frontier Economics | September 2014
Contents 14-09-03 Assessment of ARTCs Revenue
Allocation Methodology final STC
Assessment of ARTC’s Revenue
Allocation Methodology
Executive summary iii
1 Introduction 1
1.1 The Hunter Valley Coal Chain 1
1.2 The HVAU and the ACCC process 3
1.3 Structure of this report 5
2 ARTC’s revenue allocation methodology 7
2.1 We have not been provided with access to ARTC’s costing model 7
2.2 ARTC uses an “unders and overs” methodology in Z1 and Z2; and a
loss capitalisation model in Z3 8
2.3 ARTC reallocates revenue from Z1 to Z3 10
3 The ARTC methodology is not consistent with regulatory
best practice 13
3.1 The ARTC revenue allocation methodology raises questions about the
appropriate cost tests 13
3.2 The ARTC methodology is likely to create incentives for inefficient
investment decisions 17
3.3 The ARTC methodology is inconsistent with the efficient use of
infrastructure 18
3.4 ARTC’s approach discriminates against users in Zones 1 and 2 in a way
that does not aid efficiency 19
3.5 The ARTC methodology is not in the interests of persons who might
want to access 20
3.6 The ARTC methodology is not necessary to meet its legitimate business
interests 20
ii Frontier Economics | September 2014
Tables and figures
Assessment of ARTC’s Revenue
Allocation Methodology
Boxes
Box 1: Stylised illustration of notional ARTC revenue allocation methodology 9
Box 2: Stylised illustration of ARTC revenue allocation methodology 11
Figures
Figure 1: Key features of the ARTC HVRN 3
September 2014 | Frontier Economics iii
Executive summary
Executive summary
Hunter Valley Energy Coal has asked Frontier Economics (Frontier) to prepare a
report that:
Reviews the Australian Rail and Track Corporation’s (ARTC’s) existing
methodology for allocating revenue between different rail zones in the
Hunter Valley
Considers whether this methodology is consistent with regulatory best
practice principles.
The Hunter Valley rail network is used by miners to transport coal from the
Hunter region to export port terminal facilities at the Port of Newcastle.
For the purposes of determining access charges to the Hunter Valley rail
network, ARTC defines three zones in its network – Zones 1, 2 and 3. When it
comes to transporting coal to the Port of Newcastle, there are three categories of
coal producer:
Those located within Zone 1, who use access to the Z1 line to transport coal
to port
Those located within Zone 2, who use access to both the Z2 and Z1 lines to
transport coal to port
Those located within Zone 3, who use access to both the Z3 and Z1 lines to
transport coal to port.
It follows, therefore, that all coal producers using the rail network transport coal
over the Z1 line; while some will also transport coal over either the Z2 or Z3
line.
Our review of ARTC’s existing costing and revenue allocation methodology has
been hampered, to some extent, by our inability to view the actual model used by
ARTC to estimate fees it charges to users of the Hunter Valley rail network. We
would recommend that this model be disclosed in the future, subject to
appropriate confidentiality arrangements, to enable users of the network to better
understand and comment on the prices they are asked to pay for access to the rail
network.
Based on our review of material that is publicly available, it would appear that
access charges set for users in Zone 3 do not cover all of the incremental costs
that they are likely to impose on the Hunter Valley rail network in the long run.
This is because Zone 3 users impose additional costs on Zone 1 of the network
other than simply the direct operating and maintenance costs they generate in
that zone. This is principally due to additional investments needed in the network
in Zone 1 to expand capacity in response to growing demand in Zone 3.
iv Frontier Economics | September 2014
Executive summary
As a result of this, payments made to ARTC by users in Zones 1 and 2 have the
effect of “cross-subsidising” payments made by users in Zone 3. In turn, this has
a number of detrimental effects, including that ARTC’s revenue allocation
methodology is:
Likely to lead to inefficient investment in both railway infrastructure and coal
mines in the Hunter Valley
Likely to lead to the inefficient use of railway infrastructure in the Hunter
Valley
Unnecessary in order to meet ARTC’s legitimate business interests
Discriminates in favour of both future and current rail users in Zone 3 in a
way that is unlikely to aid efficiency
Not in the interests of rail users in Zones 1 and 2.
September 2014 | Frontier Economics 1
Introduction
1 Introduction
1 Hunter Valley Energy Coal has asked Frontier Economics (Frontier) to prepare a
report that:
Reviews the Australian Rail and Track Corporation’s (ARTC’s) existing
methodology for allocating revenue between different rail zones in the
Hunter Valley
Considers whether this methodology is consistent with regulatory best
practice principles.
1.1 The Hunter Valley Coal Chain
2 The Hunter Valley Coal Chain (HVCC) is the supply chain for coal delivery that
links predominantly open-cut coal mines in the Hunter region in New South
Wales with export port terminal facilities at the Port of Newcastle. It also links to
domestic coal-powered fire stations in the Hunter Valley.
3 A key element in the HVCC is the existence and operation of a railway network
that is used, amongst other things, to transport coal from mines in the Hunter
region to the Port of Newcastle.1 The railway line is comprised of three main
segments:
A line that runs from the Port of Newcastle to a junction point at
Muswellbrook. For the purposes of this report, this is referred to as the
“Zone 1” (or Z1) line
A line that runs from Muswellbrook in a westerly direction toward Ulan. For
the purposes of this report, this is referred to as the “Zone 2” (or Z2) line
A line that runs from Muswellbrook in a more north-westerly direction
toward Gunnedah. For the purposes of this report, this is referred to as the
“Zone 3” (or Z3) line.
4 The Hunter Valley rail network is managed by the ARTC. The ARTC is a
Commonwealth Government-owned entity that leases the Hunter Valley rail
network from the New South Wales government under a 60-year lease that was
granted on 5 September 2004.2
1 The ACCC notes “The network is also used by non-coal traffic, including general and bulk freight
services (such as grain) and passenger services.” See ACCC, Australian Rail Track Corporation’s Hunter
Valley Rail Network Access Undertaking – Revenue allocation review – Discussion Paper, 29 May 2014, at p.
5.
2 ACCC, op. cit. at p. 5.
2 Frontier Economics | September 2014
Introduction
5 The ARTC is vertically separated and provides only a “below rail” service. We
are advised there are 4 “above rail” haulage providers operating on the Hunter
Valley rail network; and 11 coal producers operating from approximately 35 coal
mines.
6 When it comes to transporting coal to the Port of Newcastle, there are three
categories of coal producer:
Those located within Zone 1, who use access to the Z1 line to transport coal
to port
Those located within Zone 2, who use access to both the Z2 and Z1 lines to
transport coal to port
Those located within Zone 3, who use access to both the Z3 and Z1 lines to
transport coal to port.
7 It follows, therefore, that all coal producers using the rail network transport coal
over the Z1 line; while some will also transport coal over either the Z2 or Z3
line.
8 Material provided to us by Hunter Valley Energy Coal indicates that:
Z1 is the most heavily used line, with a capacity to run between 68 and 82
paths per day. It is also the shortest of the three lines, with a length of
approximately 100kms. The ACCC also notes that due to increasing coal
volumes since 2000, the Z1 line is now capacity constrained.3
Z2 is the second-most heavily used line, with a capacity to run approximately
21 paths per day. We are also advised that the township at Ulan is
approximately 276kms from the Port of Newcastle, and that the Z2 line is
also capacity constrained.
Z3 has historically been the least-used line, and has a capacity of 14 train
paths per day. It is also the longest rail line, with the township of Gunnedah
being approximately 364 from the Port of Newcastle.
9 Importantly, it would appear that while Z1 is presently capacity constrained, this
has been partly caused by growth in the volumes of coal transported from Z3 in
recent years. Further, future growth is expected in Z3 that will necessitate further
investment in capacity on the Z1 line in order to ease congestion on this line. In
this regard, the ACCC notes that:
ARTC notes that while the heaviest coal volumes are currently at the lower end of
the Hunter Valley … the expected growth in coal mining is along the Gunnedah
Basin which is producing high rates of growth in percentage terms.
3 ACCC, op. cit, at p. 6.
September 2014 | Frontier Economics 3
Introduction
The increase in coal volumes has necessitated investment in the network, in part to
accommodate the expanding volumes in PZ3. In its 2013 Hunter Valley Corridor
Capacity Strategy ARTC noted that ‘coal demand on the (Gunnedah Basin) line has
already increased significantly and is forecast to increase very rapidly. Considerable
increase in capacity continues to be needed to accommodate this growth.4
10 This suggests that recent and expected growth in coal volumes in Zone 3 are
pushing additional cost onto other users of the rail network in Zone 1. This is
both in terms of increased capacity constraints, and the need for further capital
investments that will need to be recovered from all users of Zone 1 in the long-
term.
11 Figure 1 below sets out our understanding of some of the key features of the
Hunter Valley rail network.
Figure 1: Key features of the ARTC HVRN
Source: Frontier Economics
1.2 The HVAU and the ACCC process
12 The terms and conditions under which the ARTC supplies access to its below rail
services in the Hunter Valley rail network are governed by the HVAU, which was
approved by the ACCC in June 2011. In determining whether to accept the
undertaking, the ACCC was required to have regard to a number of matters set
out in section 44ZZA(3) of the Competition and Consumer Act (CCA),
including:
the objects of Part IIIA of the CCA, which are to:
4 ACCC, op. cit., at p. 6.
4 Frontier Economics | September 2014
Introduction
promote the economically efficient operation of, use of and investment in
the infrastructure by which services are provided, thereby promoting
effective competition in upstream and downstream markets; and
provide a framework and guiding principles to encourage a consistent
approach to access regulation in each industry.
the legitimate business interests of the provider of the service
the public interest, including the public interest in having competition in
markets (whether or not in Australia)
the interests of persons who might want access to the service
any other matters that the ACCC thinks are relevant.
13 In addition to these matters, the ACCC was also required to have regard to
pricing principles specified in section 44ZZCA of the CCA, which provide that:
regulated access prices should:
be set so as to generate expected revenue for a regulated services that is at
least sufficient to meet the efficient costs of providing access to the
regulated service; and
include a return on investment commensurate with the regulatory and
commercial risks involved; and
access price structures should:
allow multi-part pricing and price discrimination when it aids efficiency;
and
not allow a vertically integrated access provider to set terms and
conditions that discriminate in favour of its downstream operations,
except to the extent that the cost of providing access to other operators is
higher; and
access pricing regimes should provide incentives to reduce costs or otherwise
improve productivity.
14 The ACCC has recently released a Discussion Paper describing the approach
used by the ARTC to determine access prices for users in the different zones in
the Hunter Valley rail network. This involves what amounts to a “reallocation” of
revenue that would otherwise be collected from users in Zone 3 for their use of
the rail line in Zone 1 such that it is recovered from other users in Zones 1 and 2.
15 The ACCC has asked interested parties to comment on any matters relevant to
the ARTC’s revenue allocation approach.
September 2014 | Frontier Economics 5
Introduction
1.3 Structure of this report
Our report is structured so that:
Section 2 sets out our understanding of ARTC’s revenue allocation
methodology
Section 3 considers whether this methodology is consistent with the criteria
set out in section 44ZCA(3) of the CCA that are used to assess access
undertakings under Part IIIA of the CCA.
September 2014 | Frontier Economics 7
ARTC’s revenue allocation methodology
2 ARTC’s revenue allocation methodology
16 In this chapter of our report, we briefly describe the methodology used by the
ARTC to determine prices for access to lines in its rail network. Importantly, we
observe that:
The ARTC adopts a different approach to recovering its estimates of
economic cost in Z1 and Z2 compared to that which it uses to recover costs
in Z3. In particular, while it uses an “unders and overs” methodology to
recover costs in Z1 and Z2, it uses a “loss capitalisation approach” to recover
costs in Z3.
The ARTC makes additional adjustments to its approach to determine
revenues it collects from access seekers in different zones that has the effect
of reducing the amounts it will need to recover from users in Z3, and
increasing the amounts it recovers from users in Z1 and Z2.
17 It is this second “reallocation” of revenues that is the subject of the ACCC’s
Discussion Paper, and which raises concerns that we discuss in detail in section 3
of this report.
2.1 We have not been provided with access to
ARTC’s costing model
18 At the outset, it is important to note that we have been hindered in our ability to
fully analyse the approach taken by the ARTC to determine the revenues it seeks
to recover from individual access seekers. This is because we have not been
provided with access to the model it uses to determine prices paid by individual
access seekers. This creates two levels of concern:
First, in order to understand the approach it has taken to recover costs from
individual access seekers, we are reliant on the ACCC’s descriptions of the
ARTC’s approach, as set out in the ACCC’s Discussion Paper on this issue.
In this respect, we have only been provided with qualitative descriptions of
the ARTC’s approach, and have not been able to inspect the actual model it
has used. It is possible, therefore, that there are other adjustments made in
ARTC’s modelling that have material effects on its approach to setting prices
that we are simply not aware of, and that might impact on our view of its
approach.
Second, even where we do understand the method employed by ARTC to
determine the revenues it seeks to collect from individual access seekers, we
are unable to see whether it has correctly applied this method in its model.
That is, we have been unable to observe whether there are any errors in the
formulations in the ARTC model that are materially affecting the revenues it
seeks to collect from individual access seekers.
8 Frontier Economics | September 2014
ARTC’s revenue allocation methodology
19 We find this lack of transparency surprising, and inconsistent with approaches to
setting prices that we have observed for other regulated infrastructure services. In
our experience, parties are typically able to have access to – and to interrogate –
the models used by infrastructure service providers to determine the prices they
must pay for regulated services. We would strongly recommend that, subject to
appropriate confidentiality arrangements, the ARTC disclose any models and
workings it has used to determine revenues it seeks to collect from individual
access seekers for both existing and future periods.
2.2 ARTC uses an “unders and overs” methodology
in Z1 and Z2; and a loss capitalisation model in
Z3
20 Based on the ACCC’s Discussion Paper, it would appear that users of ARTC’s
below-rail service are charged a “two-part tariff”. That is, each user is charged:
A fixed fee that is meant to reflect a contribution towards the recovery of
fixed operating costs and the depreciation of (and return on) assets. This is
charged on a take-or-pay basis, and is referred to in the ACCC’s Discussion
Paper as a “FCC charge”.
A variable fee, which is meant to reflect recovery of the direct operating and
maintenance costs of providing services to it. This is charged on a non-take-
or-pay basis, and is referred to in the ACCC’s Discussion Paper as a “VCC
charge”.
21 The ARTC pricing methodology would appear to:
First, estimate amounts that it should be entitled to recover in each year that
would enable it to recover the “economic” cost of providing its services in
each zone over the lifetime of the relevant assets
Second, compare this to the revenues it collects from the two-part tariffs
referred to above
Third, ensure that the VCC it charges each user covers the direct costs of
providing services on each route to each user (this is the so-called “floor
limit”)
Fourth, compare the revenues it receives from the FCC to those amounts it
considers it is entitled to recover as contributions towards its estimated fixed
costs for that route.
September 2014 | Frontier Economics 9
ARTC’s revenue allocation methodology
22 The key issue raised in the ACCC’s Discussion Paper relates to cases where the
amounts recovered from the FCC charges for a given zone are not equal to the
amounts ARTC believes it is entitled to recover for that zone in a given year (as
determined by its financial/economic cost model). Notionally, where:
The combined FCCs initially allocated for use in Z1 and Z2 are not equal to
the combined amounts it is entitled to recover across these two zones under
the ARTC model, the difference is settled via an “unders and overs”
methodology. This means that if the combined FCCs under-recover the
combined estimated costs across these two zones, the ARTC will seek
additional revenues from users in these zones. In contrast, if the combined
FCCs over-recover combined estimates of costs in a period, the excess is
returned/refunded to users in those zones.
The FCC initially allocated for use in Z3 is not equal to the amounts it is
entitled to recover under the ARTC model, the difference is taken into
account via a “loss capitalisation model”. At present, Z3 is under-utilised and
so FCC charges are unable to recover the economic costs ARTC has
modelled for current periods. This means there is a deficit between the FCC
and the economic costs for these periods in Z3. In the case of a typical cost
allocation approach using a loss capitalisation model, this difference would be
added to the capital base that the ARTC would then be expected to recover
in future periods once Z3 reaches capacity. However, for the reasons
discussed in section 2.3 below, the ARTC does not presently adopt this
approach.
23 Box 1 below provides a stylised illustration of how the ARTC’s approach would
notionally work to determine revenues that should be collected from users in
different zones.
Box 1: Stylised illustration of notional ARTC revenue allocation methodology
Step 1 – Assume the ARTC’s economic model estimates the following amounts
should be recovered from users in each zone in a given year:
Zone VCC FCC
Z2 $10m $20m
Z3 $5m $15m
Z1 (1) $3m $8m
} $20m Z1 (2) $2m $6m
Z1 (3) $2m $6m
Nb Z1(1) refers to the revenues that should be collected from Z1 users in Z1; Z1(2) refers to
10 Frontier Economics | September 2014
ARTC’s revenue allocation methodology
the revenues that should be collected from Z2 users in Z1 etc
Step 2 – Assume the ARTC is able to collect the following revenues from each user
in each zone (nb. under its pricing methodology, its revenues must recover direct
costs (i.e. the VCC must be fully recovered)). The resultant over or under-recovery of
the FCC charge is shown in the final column.
Zone VCC FCC Over/under
recovery of FCC
Z2 $10m $22m +$2m
Z3 $5m $4m -$11m
Z1 (1) $3m $9m
} $23m +$3m5 Z1 (2) $2m $7m
Z1 (3) $2m $7m
In these circumstances, if the notional methodology were applied without the ARTC’s
second round of revenue allocations (as discussed in section 2.3 below), the
following would occur:
The combined $5m over-recovery6 in Z1 and Z2 would be would be returned to
Z1 and Z2 users under the overs and unders methodology
$11m would be added to the loss capitalisation measure in the asset base for Z3
users.
Source: Frontier Economics
2.3 ARTC reallocates revenue from Z1 to Z3
24 In practice, however, ARTC does not simply follow the method set out in
section 2.2 above. Instead, it makes further adjustments to its revenue allocation
methodology. When determining whether FCC revenues recover estimates of
economic cost, the ARTC:
Does not set FCC revenues it receives from Z3 users in Z1 against the
estimates of economic cost in Z1
Instead takes revenues from Z3 users in Z1 and sets them against the
economic costs it considers it is entitled to recover in FCC charges in Z3.
5 This is estimated as the sum of the difference between FCC revenue and cost for each user in Z1
(i.e. $9m-$8m for Z1(1) + $7m-$6m for Z1(2) + $7m-$6m for Z1(3) = $1m + $1m + $1m = $3m).
6 That is, the $2m over-recovery from Z2 plus the total $3m over-recovery from all users in Z1.
September 2014 | Frontier Economics 11
ARTC’s revenue allocation methodology
25 The effect of this is two-fold:
First, it means there is less FCC revenue in Z1 to set against ARTC’s
estimates of economic cost in that zone. This increases the amount of
revenues the ARTC will recover from Z1 and Z2 users in Z1. That is, it will
either reduce the amount of over-recovery returned to Z1 and Z2 users; or
increase the amount of under-recovery collected from them under the “overs
and unders” methodology.
Second, it will increase the FCC revenue deemed to have been recovered in
Z3. In turn, this will reduce the loss incurred in Z3, and thereby reduce the
amount added into the asset base under the loss capitalisation model. This
will have the effect of reducing the amounts of revenue the ARTC needs to
collect from Z3 users in future periods once this zone in the rail network
becomes capacity constrained.
26 A stylised example describing how the revenue allocation methodology works is
set out in the box below.
Box 2: Stylised illustration of ARTC revenue allocation methodology
The revenues notionally collected using the example in Box 1 above are repeated
immediately below.
Zone VCC FCC
Amended
Over/under
recovery of FCC
Z2 $10m $22m +$2m
Z3 $5m $4m -$11m
Z1 (1) $3m $9m
} $23m +$3m7 Z1 (2) $2m $7m
Z1 (3) $2m $7m
Rather than collect revenues in this way, however, ARTC would take the $7m FCC
charge recovered from Z3 users in Z1 (which is shaded in the table above), and
reallocate this toward the recovery of the FCC in Z3. The effect of this would be two-
fold:
it would convert a $5m combined over-recovery across Z1 and Z2 into a
combined $2m under-recovery across these zones. This would then be
7 This is estimated as the sum of the difference between FCC revenue and cost for each user in Z1
(i.e. $9m-$8m for Z1(1) + $7m-$6m for Z1(2) + $7m-$6m for Z1(3) = $1m + $1m + $1m = $3m).
12 Frontier Economics | September 2014
ARTC’s revenue allocation methodology
recovered from users in Z1 and Z2 leading to them paying more than would
otherwise be the case
it would reduce the $11m otherwise entered into the loss capitalisation account in
Z3 to only $4m for this period. This will have the effect of reducing the amounts
the ARTC would seek to recover from Z3 users in future periods once the
network becomes capacity constrained and can recover costs.
Source: Frontier Economics
September 2014 | Frontier Economics 13
The ARTC methodology is not consistent
with regulatory best practice
3 The ARTC methodology is not consistent
with regulatory best practice
27 It should be noted at the outset that we take no objection to a cost recovery
scheme which comprises both an “unders and overs” and cost allocation
approach to different components of the relevant infrastructure (as contemplated
generally by the HVAU). What we are concerned about is the manner in which
the ARTC has sought to apply the cost allocation methodology in practice.
28 In providing our assessment of the ARTC’s revenue allocation methodology, we
have assessed its approach against the criteria set out in section 44ZZCA(3) the
CCA, and referred to in section 1.2 of this report.
29 Our review of the information available on the ARTC revenue allocation
methodology leads us to conclude that its approach is:
Unlikely to ensure revenues from users in Zone 3 recover the incremental
costs imposed on the rail network by these users
Likely to lead to inefficient investment in both railway infrastructure and coal
mines in the Hunter Valley
Likely to lead to the inefficient use of railway infrastructure in the Hunter
Valley
Unnecessary to meet ARTC’s legitimate business interests
Discriminates in favour of both future and current rail users in Zone 3, and is
therefore not in the interests of rail users in Zones 1 and 2
Does not involve multi-part pricing and price discrimination in a way that is
likely to aid efficiency.
30 Each of these matters is discussed in more detail below.
3.1 The ARTC revenue allocation methodology raises
questions about the appropriate cost tests
31 The HVAU specifies that the ARTC will set access prices for users of the Hunter
Valley rail network that ensures it receives revenue that lies between floor and
ceiling limits. In this regard, section 4.2 of the HVAU specifies that:
Access revenue from every access holder must at least meet the Direct Cost
imposed by that Access Holder (the Floor Limit)
The HVAU caps the maximum amount of revenue that ARTC is entitled to
receive at the Economic Cost of providing services (the Ceiling Limit).
32 ARTC notes that the purpose of the Floor Limit:
14 Frontier Economics | September 2014
The ARTC methodology is not consistent
with regulatory best practice
… is to avoid cross-subsidisation, that is, each traffic must at least cover the costs
that would be avoided if it did not use the network.8
33 Direct Cost is defined in the HVAU to mean:
… efficient maintenance expenditure, and other costs that vary with the usage of the
network but excluding Depreciation.9
34 Separately, the HVAU also includes an objective that for each segment or group
of segments, access revenue from access holders should, as an objective, meet
the incremental cost of those segments.
3.1.1 Principles of incremental cost and cross-subsidisation
in the economics literature
35 In the economics literature, incremental costs are the additional costs that a firm
incurs in providing a service relative to it not providing that service at all. In this
regard, Faulhaber states that the:
The incremental cost of a service or group of services is the additional cost of
providing that service or group of services over and above the cost of providing all
the remaining services.10
36 Where a firm provides a service to a number of parties, the incremental cost of
providing the service to only one of these parties (say A) is the difference in total
cost between providing the service to all parties minus the total cost of providing
the service to all parties other than A. To illustrate, if the total cost of providing
the service to firms A, B and C were $100; but the cost of providing the service
only to B and C is $80, then the incremental cost of providing the service to A is
$20.
37 It is commonly understood in the literature that the incremental costs of
providing a service to one firm can be determined by calculating those costs that
would be avoided if that service were no longer provided to the firm. This is
consistent with the definition of incremental costs referred to by the ACCC in its
Discussion Paper, where it notes that:
Incremental costs are defined as all costs that could be avoided in the medium term
if a Segment was removed from the network.11
8 ARTC, Revenue Allocation Review Submission, at p. 4.
9 ACCC, op. cit., at p. 9.
10 Faulhaber, G., Cross-subsidy Analysis with more than Two Services, A note for Sprint, August 2002 at p. 1.
The note can be found at:
http://assets.wharton.upenn.edu/~faulhabe/cross%20subsidy%20analysis.pdf
11 ACCC, op. cit., at p. 9.
September 2014 | Frontier Economics 15
The ARTC methodology is not consistent
with regulatory best practice
38 This is also consistent with the approach taken by the ACCC to define
incremental costs in other regulated industries, such as telecommunications
where it notes that:
[Total service long-run incremental cost] TSLRIC is the incremental or additional
costs the firm incurs in the long term in providing the service, assuming all of its other
production activities remain unchanged. It is the cost the firm would avoid in the long
term if it ceased to provide the service.12
[emphasis added]
39 Similarly, in relation to post, the ACCC states that:
The incremental cost of a service is defined as the additional cost incurred in
producing that service (in addition to the other services the firm produces). Another
way of considering incremental cost is to ask what costs would be avoided, in the
long run, if the service were no longer offered.13
40 It follows, therefore, that the incremental costs of providing access to a rail
service to an individual user would be equal to the costs that would be avoided if
that user were no longer provided access to the service.
41 Importantly, however, it is clear that direct cost does not have the same meaning
as incremental cost. That is, while direct costs are a form of incremental cost,
they are in many cases merely a subset of incremental costs. This is because, in
the long-run, more than simply direct costs may be able to be avoided if a service
(or group of services) or a segment is no longer provided.
42 The distinction between incremental costs and direct costs is clear in other
ACCC observations, including those in relation to post where it makes clear that
incremental costs involve both direct and attributable costs:
Costs that are direct to a particular service will be incremental to that service
as they are ‘solely associated with a particular service’ and would therefore
be avoided if that service were no longer offered.
A cost that is attributable to a group of services is incremental to that
combination of services (i.e. if that combination of services were no longer
offered, the cost would be avoided) and may be incremental to a particular
individual service. The extent to which a particular attributable cost is
incremental to a particular individual service depends on the extent to which
Australia Post can avoid this particular cost by not providing that particular
service.14
43 In our view, it is clear that, in the long-run, all costs that could be avoided if users
in a rail segment were no longer provided with access to a service should fall
within the meaning of incremental costs. This would include both variable and
fixed (including depreciation) costs that would be avoided if the segment were no
12 ACCC, Access pricing principles – Telecommunications, a guide, July 1997 at p.28.
13 ACCC, Tests for assessing cross-subsidy, June 2014 at p. 5.
14 Ibid.
16 Frontier Economics | September 2014
The ARTC methodology is not consistent
with regulatory best practice
longer provided. This is an important distinction because, as noted above, the
definition of Direct Costs applied by ARTC explicitly excludes depreciation. This
would appear to include depreciation even if that relates to capital costs that
could be avoided in the medium term (or long-run) if users in a particular zone
(e.g. Zone 3) were no longer provided with a rail service.
44 It is also well understood in the economics literature that a cross-subsidy occurs
to an individual (or group of individuals) when that individual (or group of
individuals) is charged a fee that does not cover its incremental costs – and where
someone else is charged more than their stand-alone costs of providing a
common service. In this regard, a classic article in the economic literature relating
to the meaning of cross-subsidies was published by Professor Gerald Faulhaber
in 1975.15 In his paper, Professor Faulhaber considered the example of a rail
network that had added a new line to its rail network. He then uses this example
to consider under what conditions prices for the new individual (or incremental)
rail line connecting a town to the network would involve a cross-subsidy. In this
respect, he notes that:
Provided the revenues realized from providing rail service to the town exceed the
added costs, the answer must be in the negative.16
45 In a note further explaining his seminal article on cross-subsidies, Professor
Faulhaber states that:
… if the revenues of a regulated enterprise just cover total economic costs, then all
prices are subsidy-free if the revenues of each service and each group of services is
at least as great as the incremental cost of that service or group of services …17
46 In other words, if the revenue from one service (or group of services) of a multi-
product firm covers the incremental (or additional) costs of providing that
individual service (or group of services), then that service (or group of services)
can not be said to be subject to a cross-subsidy.
3.1.2 ARTC’s revenue allocation methodology appears to
involve a cross-subsidy to users in Zone 3
47 We contend that ARTC’s revenue allocation practices and application of the tests
with respect to users in Zone 3 do not ensure that no cross-subsidisation occurs.
15 Faulhaber, G., “Cross-Subsidization: Pricing in Public Enterprises”, American Economic Review, 65(5)
December 1975, pps. 966-977.
16 Ibid., at p. 966.
17 Faulhaber, G., Cross-subsidy Analysis with more than Two Services, A note for Sprint, August 2002 at p. 1.
The note can be found at:
http://assets.wharton.upenn.edu/~faulhabe/cross%20subsidy%20analysis.pdf
September 2014 | Frontier Economics 17
The ARTC methodology is not consistent
with regulatory best practice
48 This contention follows simply from the notion that the avoidable cost of serving
mines located in Zone 3 with access to track located in Zone 1 is not restricted to
the direct operating and maintenance expenses of supplying Zone 3 users in
Zone 1 if there are some capacity constraints in Zone 1. If such constraints exist,
then it implies that the avoidable costs of supplying users in Zone 3 will include
the costs of the capacity used to supply these users. In other words, if their
demand did not exist, the network in Zone 1 would have lower capital costs in
the medium term and the long run. It would only be appropriate to ignore the
costs of such capacity if there was substantial excess capacity in the network, as
then the long run costs of suppling users in Zones 1 and 2 would be no different
if Zone 3 users were excluded from using Zone 1 in the rail network.
49 Our understanding is that ARTC has been investing in further capacity in Zone 1,
and will continue to invest in further capacity in the future.18 Further, we
understand that this additional investment is being driven (at least in part) by
increasing use of Zone 1 by users located in Zone 3. These costs – which are
incremental or avoidable in relation to users located in Zone 3 – should be
recovered from those users. As it stands, under the revenue allocation
methodology it is (at best) unclear how much of the incremental cost will be
recovered from users in Zone 3. It would be inefficient for these capital costs to
be recovered by users in Zone 1 and Zone 2, as we now go on to describe.
3.2 The ARTC methodology is likely to create
incentives for inefficient investment decisions
50 In considering efficient investment, it is necessary to consider investment in both
the markets for the supply of below rail services, and in the markets up or
downstream of the railway network (in this instance, the supply of coal).
51 Efficient investment in below rail services is promoted by regulation that allows
the access provider to recover its efficient costs of supply on each route section.
Efficient investment in downstream markets requires both that costs reflect
efficient costs and that the pricing regime limits access providers’ ability to
expropriate the sunk investments of downstream firms (i.e. in mines).19
52 The current scheme of cost recovery allows the ARTC not to recover the (long
run) efficient incremental costs of supplying services in Zone 1 to mines located
in Zone 3. Users in Zones 1 and 2 currently cover all capital costs in Zone 1 –
even though the use of the zone by users in Zone 3 is adding additional capital
18 ARTC, op. cit., p. 15.
19 On the first element, see Australian Competition Tribunal, Re Telstra Corporation Ltd (No 3) [2007],
at 164. On the second element, see Darryl Biggar, Is Protecting Sunk Investments by Consumers a
Key Rationale for Natural Monopoly Regulation?, at p. 2, available at:
https://www.accc.gov.au/system/files/Darryl%20Biggar%20paper.pdf
18 Frontier Economics | September 2014
The ARTC methodology is not consistent
with regulatory best practice
costs to the build of Zone 1 of the rail network. While eventually there is an
expectation that Zone 3 users will recover some of the capital costs in Zone 120,
there appears to be no intention for these users to recover all of their incremental
costs in Zone 1.
53 This system of cost recovery seems to favour investment in Zone 3 rather than in
Zones 1 and 2, even where the marginal costs of producing coal are the same.
Indeed, the ARTC appears to suggest that this is a favourable outcome of its
scheme, as it encourages entry and hence the growth and development of new
coal basins.21
54 Economic efficiency is ultimately about maximising value. Maximising value
comes from finding the largest gap between the costs of extracting and shipping
the coal and the market price of coal. The pricing of below rail services should
facilitate the achievement of this efficiency. In contrast, it would be undesirable
to price below rail services to encourage the entry of new mines in Zone 3 at the
expense of existing mines in Zones 1 and 2, or new mines located in Zones 1 and
2.
55 Further, the revenue allocation scheme seems to offer little certainty to
downstream firms about the access provider’s efficient costs of supply on
particular route sections. Nor does it appear to effectively constrain the ARTC
from expropriation from downstream producers that are particularly successful,
in order for the ARTC to achieve other objectives such as the development of
new mines. These firms might, for example, be required to pay all capital costs
on segments in Zones 1 and 2, while firms located in Zone 3 have the advantage
that they are not required to contribute to the capital costs of Zone 1, even
where some of those costs are incremental to their use.
3.3 The ARTC methodology is inconsistent with the
efficient use of infrastructure
56 The efficient use of infrastructure is largely determined by the relationship of
charges to the marginal costs of supplying the infrastructure. That is, the most
efficient use of existing infrastructure will result where users face the marginal
cost of carrying an extra tonne of coal (or carriage, or train, depending on the
particular unit of output).
57 It appears that, in general, the ARTC pricing methodology may not hinder the
efficient use of below rail infrastructure in the short-run. This is because its
variable charges are designed to reflect variable (direct) costs. However, the
20 ARTC submission, p. 14.
21 Ibid. p. 2.
September 2014 | Frontier Economics 19
The ARTC methodology is not consistent
with regulatory best practice
current cost recovery scheme does not appear to promote the allocative
efficiency of resource use more broadly in the long-run. This is because the
ARTC’s costing and revenue allocation methods appear to involve a cross-
subsidy from Zone 1 and 2 users to users in Zone 3. In turn, this means that the
ARTC approach will encourage the extraction and transport of coal that is higher
cost – in Zone 3. ARTC explicitly recognises this when it says it negotiates a cost
of access that reflects a balance between:
promoting the development and expansion of the Gunnedah Basin... and the
recovery of a reasonable level of the cost of recent investment in PZ3...22
58 It follows, therefore, that the methodology and pricing decisions:
(a) do not maximise the economic value of the coal resources; and
(b) result in patterns of usage of the network that are not consistent with
those which would maximise economic efficiency overall.
3.4 ARTC’s approach discriminates against users in
Zones 1 and 2 in a way that does not aid
efficiency
59 It is clear from our examination of material on ARTC’s revenue allocation
methodology that its approach does discriminate in favour of Zone 3 users, and
against users in Zones 1 and 2. This is because it appears to effectively ensures
that users in Zone 3 presently make no contribution toward the recovery of the
capital costs associated with the provision of services in Zone 1 – even if these
costs could be avoided if users in Zone 3 were no longer provided access to the
rail network. It instead ensures that these costs are recovered from users located
in Zones 1 and 2.
60 It is true that the pricing principles in section 44ZZCA of the CCA allow for
multi-part pricing and price discrimination when it aids efficiency. However, for
the reasons set out in section 3.3 above, it is not our view that the form of price
discrimination employed by ARTC does aid efficiency. Indeed, where ARTC’s
revenue allocation approach leads to a cross-subsidy from users in Zones 1 and 2
to users in Zone 3, it is likely to lead to inefficiency in the use of and investment
in the rail network.
22 ARTC, op. cit., at p. 16.
20 Frontier Economics | September 2014
The ARTC methodology is not consistent
with regulatory best practice
3.5 The ARTC methodology is not in the interests of
persons who might want to access
61 The Australian Competition Tribunal (Tribunal) has previously considered the
meaning of the interests of persons who have a right to use telecommunications
access services under Part XIC of the CCA. In relation to the provision of a
particular telecommunications services (a line sharing service), the Tribunal
found that:
The interests of persons who have a right to use the LSS, access seekers, are
served by an access price that enables them to compete on their merits (that is, on
the basis of their own efficiency) in downstream markets.23
62 In our view, the revenue and costing allocation approach adopted by the ARTC
has the potential to inhibit the ability of users of the rail network to compete on
their merits in downstream markets. In this respect, firms that are less efficient at
extracting coal and transporting it to export markets should not be able to
compete and survive in downstream markets on the basis of any price
discrimination with respect to the provision of access to railway services. The
pricing approach adopted by the ARTC does, however, have the potential to
allow this to occur. This is because it involves coal miners in Zones 1 and 2
cross-subsidising coal miners in Zone 3. It also has the effect of raising the
marginal costs of users in Zones 1 and 2 relative to those in Zone 3 even when
users in Zones 1 and 2 may not impose any greater marginal costs on ARTC than
those in Zone 3.
63 In our view, therefore, it is not in the interests of users (i.e. persons) in Zones 1
and 2 that do not operate in Zone 3 to either:
Cover in total the capital costs involved in the provision of rail services in
Zone 1 in a way that ensures users in Zone 3 make no contribution toward
the capital costs of Zone 1
Provide a cross-subsidy to users in Zone 3 by footing the bill for investments
to increase the capacity of Zone 1 to accommodate use of the rail network in
that zone by users in Zone 3.
3.6 The ARTC methodology is not necessary to meet
its legitimate business interests
64 The Tribunal has, on a number of occasions, had cause to consider what is meant
by “the legitimate business interests” of an access provider in the context of its
assessment of a number of telecommunications access price undertakings
23 Re: Telstra Corporation Limited (2006) ATPR 42-121, at para 138.
September 2014 | Frontier Economics 21
The ARTC methodology is not consistent
with regulatory best practice
provided by Telstra. In this regard, the Tribunal found in Telstra Corporation
Limited [2006] at para [89] that:
… legitimate business interests require that Telstra be allowed to recover its costs of
supplying …[a service] … and achieve a normal return on its invested capital …. It is
a reference to the interest of a carrier in recovering the costs of its infrastructure and
its operating costs and obtaining a normal return on its capital.
65 We accept that ARTC should be entitled to recover the costs of its investment in
its rail network (inclusive of a normal return on its capital investments) where
market conditions allow this to occur.
66 We also accept that ARTC’s revenue allocation methodology is not intended to
ensure it is able to recover more than its costs (inclusive of a normal return on its
investment). In this respect, the revenue allocation methodology would appear to
simply reallocate the recovery of capital costs between different users of its rail
network.
67 We would not accept, however, that ARTC’s revenue allocation methodology is
necessary for it to meet its legitimate business interests. The effect of its revenue
allocation methodology appears to be to reduce the size of its capitalised losses in
Zone 3 in current periods while increasing the amount of capital costs it recovers
from users in Zones 1 and 2. While such a method will increase the speed with
which ARTC is able to recover its costs, it is not in our view necessary to ensure
it is able to recover its costs in the long-run.
68 In our view, the loss capitalisation model (LCM) established for Zone 3 provides
an adequate measure that enables ARTC to recover, over time, its costs of
providing services in Zone 3 (and indeed across all three zones in its network).
The issue here is essentially one of timing, in that the current method allows for a
smaller LCM and a faster recovery of the economic costs of the entire network.
However, it is equally available to ARTC to recover the efficient level of costs
from Zone 3 users via the LCM. To the extent that the LCM does not guarantee
cost recovery in Zone 3, then this is a risk that the infrastructure owner should
be prepared to bear – not users in Zones 1 and 2.
69 Further, it is not clear to us that ARTC’s legitimate business interests extend to it
recovering the costs of infrastructure from Zone 1 users via cross subsidies to
users in Zone 3 (to the extent that Zone 3 users do not recover the incremental
costs of their use in Zone 1).
Frontier Economics Pty Ltd in Australia is a member of the Frontier Economics network, which
consists of separate companies based in Australia (Melbourne & Sydney) and Europe (Brussels,
Cologne, Dublin, London & Madrid). The companies are independently owned, and legal
commitments entered into by any one company do not impose any obligations on other companies
in the network. All views expressed in this document are the views of Frontier Economics Pty Ltd.
Disclaimer
None of Frontier Economics Pty Ltd (including the directors and employees) make any representation or warranty as to the accuracy or completeness of this report. Nor shall they have any liability (whether arising from negligence or otherwise) for any representations (express or implied) or information contained in, or for any omissions from, the report or any written or oral communications transmitted in the course of the project.
FRONTIER ECONOMICS | MELBOURNE | SYDNEY
Frontier Economics Pty Ltd 395 Collins Street Melbourne Victoria 3000
Tel: +61 (0)3 9620 4488 Fax: +61 (0)3 9620 4499 www.frontier-economics.com
ACN: 087 553 124 ABN: 13 087 553 124